China's Missing Pigs: Correcting China's Hog Inventory Data Using a Machine Learning Approach
Yongtong Shao is a professor in the Department of Finance at the Tianjin University of Commerce, 409 Guangrong Rd, Beichen District, Tianjin, China. Tao Xiong is a professor and chair in the Department of Agricultural Economics & Management at Huazhong Agricultural University, 1 Shizhishan Rd, Nanhu District, Wuhan, China. Minghao Li is an assistant professor in the Department of Economics, Applied Statistics & International Business at New Mexico State University, Domenici Hall 212, Las Cruces, NM. Dermot Hayes is Charles F. Curtiss Distinguished Professor in Agriculture and Life Sciences in the Department of Economics, the Department of Finance, and the Center for Agricultural and Rural Development at Iowa State University, 518 Farmhouse Lane, 568C Heady Hall, Ames, IA 50011. Wendong Zhang is an assistant professor in the Department of Economics and the Center for Agricultural and Rural Development at Iowa State University, 518 Farmhouse Lane, 478C Heady Hall, Ames, IA. Wei Xie is a graduate student in the Department of Finance at the Tianjin University of Commerce, 409 Guangrong Rd, Beichen District, Tianjin, China. Li, Zhang and Hayes gratefully acknowledge support from the USDA National Institute of Food and Agriculture Hatch Project 101,030 and grant 2019‐67023‐29414, and Xiong acknowledges the support from the National Natural Science Foundation of China (Project No. 71771101). The authors thank the ISU Center for China‐US Agricultural Economics and Policy, where Li was a postdoctoral research associate and Shao and Xiong were visiting scholars. The authors also appreciate editing assistance from Nathan Cook, Becky Olson, and Barbara Nordin, and comments from Guiping Hu, Chad Hart, and Kelvin Leibold. Any remaining errors are the authors' responsibility. Wendong Zhang is the corresponding author, and Yongtong Shao and Tao Xiong are jointly first authors. For questions about the online replication codes, please contact Xiong.
Abstract
Small sample size often limits forecasting tasks such as the prediction of production, yield, and consumption of agricultural products. Machine learning offers an appealing alternative to traditional forecasting methods. In particular, support vector regression has superior forecasting performance in small sample applications. In this article, we introduce support vector regression via an application to China's hog market. Since 2014, China's hog inventory data has experienced an abnormal decline that contradicts price and consumption trends. We use support vector regression to predict the true inventory based on the price‐inventory relationship before 2014. We show that, in this application with a small sample size, support vector regression outperforms neural networks, random forest, and linear regression. Predicted hog inventory decreased by 3.9% from November 2013 to September 2017, instead of the 25.4% decrease in the reported data.
